The long-term goal of this project is to elucidate the cognitive mechanisms implicated in numerical processing deficits (dyscalculia) resulting from brain damage. The proposed studies have three specific aims. The first is to test hypotheses embodied in a model that specifies the functional architecture of the cognitive numerical processing system, and interprets the major forms of dyscalculia in terms of functional damage to particular components of the system. The second aim is to analyze in detail the internal structure of individual cognitive components within the numerical processing system, thereby providing a basis for a more finegrained analysis of the wide variety of acquired numeral-processing and calculation deficits. The third specific aim is to assess the extent to which the cognitive mechanisms underlying numerical processing, and the deficits resulting from disruption of these mechanisms, are number-specific or instead more general. These aims will be accomplished through a twophase research strategy combining broad-based testing of large numbers of patients, and indepth individual-patient analyses for subgroups of patients with impairment to particular processing mechanisms (e.g., verbal numeral production deficit, arithmetic fact retrieval deficit). Interacting with the empirical research program is a program of computational modeling, the aim of which is to formulate theoretical proposals about numerical processing mechanisms and their disruption in computationally explicit terms. Finally, the theoretically-grounded characterizations of functional deficits emerging from the studies, in conjunction with lesion localization data from magnetic resonance imaging, will be used to examine correlations between functional deficits and brain lesions. The findings of the proposed studies will contribute significantly to understanding of numerical processing and dyscalculia, and may also provide a foundation for development of assessment and rehabilitation methods. Finally, the results may have relevance for issues extending beyond the domain of numbers, including issues concerning the structure and dissolution of lexical processing mechanisms.